As the original article is not online, I have not read it. After reading abstract I have a suggestion. It would be more helpful if 95% confidence intervals (CI) are given with each percentage. It is very easy to calculate and usually tells more about the sample. For example, 95% CI of rheumatic heart disease is 30.24-95%CI-57.76. Age should be given with standard deviation to get some idea of the age mix.
Posted by: rqayyum - Jan 2005